Research Of Audio Watermarking Algorithm Based On Complex Cepstrum Transform

Digital watermarking is a very important technology in today’s information age.It plays an important role in the aspect of information, and is widely used in copyright protection of digital information.However, audio watermarking is more difficult to realize than others of digital watermarking technologies.This paper structures an audio watermarking algorithm based on complex cepstrum transform and an audio watermarking algorithm embedded robust and fragile watermark simultaneously based discrete wavelet transform and complex cepstrum transform. The first algorithm uses Arnold transform to scramble a binary watermarking image; selects audio sections with larger energy of time domain as positions embedded watermarking image. Then, through the quantization theory, the algorithm embeds the watermarking sequence in complex cepstrum coefficients, and the process of extracting watermark realizes blind extracting. The second algorithm uses chaos sequence generated from Logistic mapping to encrypt two original watermarking images; then, embeds the robust and fragile watermark sequence respectively in complex cepstrum coefficients of the low frequency component and the high frequency component after discrete wavelet transform using the quantization theory. The experimental results show that, the embedded watermarking image by this two means has a good imperceptibility, and has a strong robustness to some common signal processing means, such as re-sampling, re-quantization, low pass filtering,adding noise and so on. And the fragile watermark has a good sensibility to some common signal processing means.